Several influential theoretical approaches to developmental errors maintain that such errors are in part due to processing difficulties (MacWhinney, 1987; Marcus, Pinker, Uliman, Hollander, Rosen and Xu, 1992). The proposed research focuses on 1Psg subject pronoun case errors, and explores how these errors contribute to processing difficulty in sentence production. The proposed research will test two hypotheses. (A) Early in the acquisition of the 1Psg subject pronoun, there is a great deal of competition between 1 Psg word-forms vying for the subject role, which in turn, increases the processing burden in producing a sentence. (B) For some children, 1Psg subject pronoun case errors may become entrenched, that is, an incorrect form, me or my, gains superiority over the correct form I, and remains in this superior position for a long period of time (Rispoli 1998a, etc). When entrenchment occurs there is little competition taking place, reducing the burden on processing. This study uses an innovative technique for measuring processing burden in spontaneous speech, fluency rate, pioneered in Rispoli and Hadley (2001). These hypotheses will be tested in a longitudinal design, with 24 child participants using language samples of child-caregiver interaction taken at three month intervals, beginning at 2;0 and ending at 3;0. If hypothesis A is correct, we should see an increase in the fluency of sentences with 1Psg subjects over time that is not attributable to general increases in fluency. If hypothesis B is correct, deviation from the trend expected by hypothesis A should be explicable by the child's degree of entrenchment in producing the 1Psg pronoun case errors. The deviation has a predicted direction; high error rates should be accompanied by increased fluency. That is, entrenchment contributes to the sacrificing of grammatical accuracy for increased fluency.